Study on Key Drivers and Collaborative Management Strategies for Construction and Demolition Waste Utilization in New Urban District Development: From a Social Network Perspective

Construction and demolition waste (C&DW) during new district development shows the characteristics of large quantity, concentrated distribution, and long duration on both supply and demand sides. The construction of the new district has objective conditions to promote the scale development of recycling industries and achieve green development through local digestion, so the recycling indicators for the new district proposed by the government are generally higher than those of other regions. However, many new districts have yet to systematically identify key drivers (KDs) of recycling and form win-win governance mechanisms with multiple government and market subjects, resulting in uncontrolled accumulation or high-cost discard of C&DW. This paper identifies 29 recycling drivers and 8 governing subjects through literature research and a field study of five national new districts in China. Then, a 2-mode social network and two 1-mode social networks are constructed to analyze the complex interactions between drivers and governing subjects, taking Nanjing Jiangbei New District as an example. The results of the study show that most of the drivers need at least 2 governance subjects to promote together, which indicates that it is necessary to build a collaborative governance mechanism of multiple subjects. This study provides a structured framework to analyze the drivers of recycling in new district development and the collaborative governance of multiple subjects, which can provide a basis for promoting efficient recycling of new district development.


Introduction
Urban new districts, including national new districts and new districts set up by local governments, are one of the forms of high-quality and rapid urban development and expansion in the urbanization process [1]. Tere are 19 national-level new districts in China, with a total scale of more than 50,000 square kilometers, which have become an important growth pole for national and local economic and social sustainable development. During the development and construction of a new district, a large number of buildings and structures need to be demolished and constructed, thus generating a large amount of C & DW. At the same time, the development of the new district also provides a better platform for low-cost (large-scale disposal, production and sales, and short transportation distances) and high-efciency utilization of C & DW. Tis is because of the large volume and concentrated distribution of C&DW generated in the construction of new districts, the long duration of new projects, the high positioning of construction standards and environmental requirements, and the need for a large amount of new construction materials. In addition, new projects with large volumes, long durations, and large construction can also provide sites for the construction of temporary fxed resource-based disposal facilities (on-site treatment facilities that can exist for 3 to 6 years), ofering the possibility of building mobile disposal facilities, temporary fxed and long-term fxed disposal facilities that can develop in parallel with efcient resource-based disposal modes in new areas.
Based on the pressure and requirements of sustainable development, C&DW's recycling research is receiving attention not only in China, but also around the world. Many scholars have studied the drivers of recycling, emphasizing the need to explore a collaborative governance mechanism led by the government, operated by social capital, and participated by the public, and to use this collaborative efect to improve the level of recycling, so as to get out of the governance dilemma of C&DW [2][3][4]. However, few studies have systematically considered the interactions among the various drivers and the role of each governance subject on the drivers, leading to deviations between theoretical studies and actual situations. Moreover, there are no studies related to the multiple and complex relationships between recycling drivers and governance subjects that characterize the development of new districts.
In order to address the above issues, the following actions are critical: the frst is to identify the drivers afecting the recycling of new district development and the collaborative governance subjects of the whole recycling process. Ten, analyze the complex relationships among the drivers, the governance subjects, and between the drivers and the governance subjects. Te advantage of SNA over other research methods is that it can provide quantitative metrics for analyzing the relationships between nodes in a network. Terefore, this paper proposes to study the abovementioned multiple complex relationships based on SNA to provide a framework for the analysis of the new district C&DW recycling. Te results of the study can prompt multiple governance subjects to adopt efective strategies for efcient collaboration and quick linkage to achieve the new district C&DW recycling goal.
Te remainder of the paper is organized as follows: the second section provides a detailed literature review of drivers and collaborative governance afecting C&DW recycling to identify knowledge gaps in the feld. Te third section describes the research methodology. In the fourth part, the SNA model is established using 29 drivers and 8 governance subjects identifed from the literature study and the research of 5 national new districts in China as network nodes. Te SNA method is used to analyze the drivers and the characteristic indicators of the 2-mode network of governance subjects to determine the role of diferent governance subjects in drivers and to verify the relationship between drivers and governance subjects. Ten, the two 1-mode networks of drivers and governance subjects are analyzed to further reveal their internal relationships. Finally, a collaborative governance strategy is proposed according to the results of the SNA.

Literature Review
2.1. Literature on C&DW Recycling Drivers. Based on the fact that the attitudes and behaviors of owners and contractors at both the supply and demand ends of the C&DW reverse supply chain are the fundamental drivers of recycling, related research is conducted mainly from the perspective of owners and contractors. Jain et al. [5] propose that the behavioral choices of owners are mainly infuenced by perceived benefts, perceived costs, attitudes, subjective norms and perceived behavioral control, institutional pressures, and environmental awareness. Al-Sari et al. [6] and Yu et al. [7], on the other hand, suggest that the factors infuencing contractors' behavior in purchasing recycled products are mainly government measures, perceived behavioral control, perceived efectiveness, subjective norms, and the infuence of recycled product information. In the absence of a regulatory framework, the voluntary attitudes and behaviors of contractors are driven primarily by direct economic considerations.
Early research on the environmental conditions required for recycling was mainly from a policy-making perspective. Jimenez-Rivero and Garcia-Navarro [8] pointed out that government regulatory instruments must be used in conjunction with economic instruments or other control strategies, emphasizing that policies as well as relevant laws and regulations are the KDs of recycling. Jia et al. [9] further used a system dynamics model to demonstrate that a synergistic policy scheme of subsidies, landfll fees, and penalties is benefcial in promoting recycling by introducing a subsidy and penalty mechanism. With the further development of research and practice, more scholars have conducted a more systematic study of drivers by constructing analytical frameworks and models. Lopez Ruiz et al. [10] constructed an analytical framework based on circular economy theory and obtained that insufcient supply of C&DW, imperfect collection and classifcation standards, lack of market demand for recycled products, immature recycling technology, low cost of natural aggregates, poor environmental awareness, imperfect regulations and legislation, insufcient supervision by regulatory authorities, lack of incentive policies, and insufcient publicity and education on recycling are the constraints of C&DW recycling. After systematically reviewing the problems of recycling, Kabirifar et al. [11] focused on the factors associated with the recycling process, including the application and monitoring of construction site waste management programs, on-site and of-site sorting techniques, incentives or penalties for contractors, demolition techniques, and carrying out plans as recycling drivers. Liu et al. [12] identifed and quantitatively analyzed the KDs of recycling in China with the help of stakeholder theory and grey DEMATEL theory, including laws and regulations, the degree of compulsion of normative standards, environmental awareness, C&DW supply, and sales of recycled products. By reviewing 105 pieces of literature, Zhao [13] summarized 35 drivers, including the perfection of laws and regulations, the supervision of the government and the public, and the attitude of stakeholders.

Literature on C&DW Recycling Collaborative Governance
Strategy. Multisubject collaborative governance refers to the act of optimizing structure, coordinating interests, and deeply integrating knowledge and resources among multiple subjects in order to obtain complementary advantages and the nonlinear utility of system superposition [14]. Te government is the frst responsible subject and should play a leading role in the recycling process. It is also a consensus among scholars that owners, contractors, and recycling companies, which are at the supply and demand ends of the C&DW reverse supply chain, should also play an important role in the supply chain [15,16]. Liu et al. [17] pointed out that government plays an important role in the decisionmaking of recycling companies, explored the impact of two recycling capacity expansion strategies on profts and the environment, and suggested that despite their conficting goals in recycling, both government and recycling companies tend to choose the same strategy when recycling handling capacity is low. Tilaye and van Dijk [18] stated that contractors can meet owners' requirements for recycling governance by providing and following an on-site management plan that specifcally includes a clear estimate of C&DW and a comprehensive plan for collection classifcation. Liu et al. [12] proposed that recycling enterprises in the industry chain cannot rely on government subsidies alone to reduce costs, they should form a synergistic linkage with more responsible subjects upstream and downstream of the industry chain to create a new industry chain development mode. Other scholars have studied the game relationship among the government, contractors, and recycling frms, revealing the underlying mechanisms of the strategic choices of the three parties. Teir fndings contribute to the understanding of the behaviors, demands, and the role of collaborative governance of resource utilization stakeholders [19,20].

Knowledge Gap.
Existing studies on recycling drivers and multisubject collaborative governance have been conducted at some depth, but there lacks a systematic study on multiple and complex relationships between subjects and drivers, and fail to clearly reveal the best way for subjects to collaboratively govern C&DW and how to do so. Te distinctive features of new district development make the related research of recycling diferent from general areas. Terefore, the knowledge gap can be further determined.
(1) Few studies have analyzed KDs for the recycling of the new district development. Te new district development recycling governance is a more demanding and complex process. Terefore, attention to its KDs is essential to determine the best way of collaborative governance to promote efcient recycling in the new district development.
(2) A social network-based perspective would be more useful than other approaches. How to construct an analytical model that can analyze the relationship not only between drivers, but also between governance subjects and between governance subjects and drivers, and explore the ways of these interdependencies and how to adopt appropriate governance strategies to reach the new district development recycling goals is a problem to be solved.

Methodology
Social networks are classifed into unimodal and bimodal networks based on the characteristics of the set of actors and the properties of the connections between them. A bimodal network indicates that the network under the study is bimodal, i.e., it includes a metric for two sets of actors or a set of actors and a set of events [21]. Te development of C&DW in the new district involves multiple governance subjects and requires a variety of drivers to work together in a complex social network environment, and the relationship network formed by various drivers with the interaction of subjects is a 2-mode social network. Te network can present the collaborative relationship between subjects through KDs jointly promoted or infuenced by subjects. Based on this, this paper adopts a 2-mode social network for subject-driver relationship analysis and a 1-mode social network for the relationship analysis within the governing subjects and between drivers. Te specifc steps and methods are as follows: Step 1. Identifying the drivers and related governing subjects of the new district recycling.
Although there are some diferences in the composition and drivers of new district development recycling in countries with diferent systems, socioeconomic cultures, and natural environmental conditions, the basic recycling process is generally the same. It is possible to frst analyze the recycling process and the governance subjects and basic relationships involved in each stage, as shown in Figure 1. Te initial list is constructed by categorizing and organizing the composition of governance subjects and drivers analyzed from the literature, then revised and supplemented by a semistructured questionnaire survey of several comparable new districts to identify recycling governance subjects and drivers.
Step 2. Constructing a recycling governance subject-driver adjacency matrix.
Based on the identifed governance subjects and drivers, a subject-driver relationship questionnaire is developed for the assessment of the relationship between governance subjects and drivers, and the questionnaire is designed in the form of a structured matrix. In order to minimize the infuence of the subjective preferences of the dominant participants, and considering that the judging requires more specialized theoretical knowledge and the number of relationships to be judged is large, the questionnaire survey should not be conducted by distributing a large number of questionnaires, and the Delphi method is appropriate for judging relationships. Te relationship is assessed using a fve-point Likert scale, with "7" indicating very strong infuence, "5" indicating infuence, "3" indicating average infuence, "1" indicating very little infuence, and "0" indicating no infuence. After obtaining the questionnaire data, the subject-driver adjacency matrix is constructed. Te criteria for determining the relationship between the nodes of the adjacency matrix are as follows: ① when the total number of people choosing "7" and "5" is greater than half of the number of researchers, it means that the subject has infuence on drivers, and the relationship value in the matrix is set to 1. ② When the total number of people choosing "0" and "1" is greater than half of the research population, it means that the subject has no infuence on drivers, and the relationship value is set to 0. If there are situations other than ① and ② in the fnal research data, a second round of survey will be set for the specifc question data.
Step 3. Constructing a 2-mode social network model.
Te subject-driver adjacency matrix is substituted with UCINET software for model visualization, and four metrics of the 2-mode social network are analyzed: degree centrality, mediator centrality, proximity centrality, and coreperiphery analysis. Degree centrality refects the centrality of nodes, the higher the degree centrality, the higher the    associated points and the more important the position in the network; the intermediary centrality refects the interrelationship between two nonadjacent nodes in the network and the degree of dependence on other nodes in the network; the proximity centrality indicates the proximity of actors to other actors in the network; and the core-periphery structure is an ideal structural pattern that divides both rows and columns into two categories, with subjects and elements closely related in the core part and more sparsely related in the edge part.
Step 4. Proposing a new district development and recycling the collaborative governance strategy. Based on the analysis results and connotations of collaborative governance from the 2-mode social network and the two 1-mode social networks, the collaborative governance strategy is systematically proposed in two dimensions: subject-organizational collaboration and resource collaboration. Organizational collaboration is mainly refected in the positioning of roles, power and responsibility, beneft and risk distribution among governance subjects in the whole process of recycling, which is the external refection of various related relationships formed by mutual cooperation among diferent governance subjects and is the basis of the composition of the governance system. Resource collaboration is the integration and interaction of tangible and intangible resources among governance subjects and is the guarantee of the operation of the governance system. Organizational and resource collaboration is coupled through interactive relationships and mutual support, which work together to ensure the achievement of dynamic collaborative governance goals in economic, ecological, and social contexts during diferent development periods of the new area.

Network Node Determination.
Based on the establishment time, scale, and development stage of 19 national-level new districts in China, Chongqing Liangjiang New District, Guizhou Gui'an New District, Dalian Jinpu New District, Sichuan Tianfu New District, and Nanjing Jiangbei New District, which have certain comparability, are selected as the research's new subjects. Te semistructured questionnaire analysis of the new research area is used to determine the governance subjects and drivers. Te respondents to the questionnaire are mainly the government, owners, contractors (including demolition companies), transportation companies, recycling companies, and other related management personnel. A total of 60 semistructured questionnaires were flled out independently by relevant management personnel in the fve research areas after the symposium with our research team to guarantee the questionnaire quality. After statistical analysis of the questionnaire, eight types of governance subjects, government (S1), owners (S2), contractors and or demolition companies (S3), recycling companies (S4), transportation companies (S5), related R&D units (S6), fnancial institutions (S7), and the public and media (S8), are identifed as the subject nodes in the 2-mode network. 29 recycling drivers are event nodes in the 2-mode network, as shown in Table 1.

Modeling Data Acquisition.
Te subject-driver relationship questionnaire is designed as a structured matrix to assess the relationship between subjects and drivers in Nanjing Jiangbei New District. According to the set of questionnaires, 2 rounds of survey interviews were conducted. In order to improve the reliability and representability of the results, a total of 48 experts were interviewed, all of them are management personnel who have engaged in or plan to engage in the work related to the new district recycling and experts, scholars, the media, and the public who know more about the new district, as shown in the following table.

2-Mode Network Visualization and Data Analysis.
Te recycling subject-driver network consists of 8 governing subjects and 29 drivers, connected by 136 links. Te red circles represent the governing subjects, and the blue squares represent the drivers. Te links between subjects and drivers refer to the subjects' ability to handle, control, and drive the drivers. Te visualization model is shown in Figure 2.
Te original 2-mode network is converted into a 1-mode network of governing subjects ( Figure 3) and a 1-mode network of drivers ( Figure 4). In the subject-subject network shown in Figure 3, each node represents a type of subject, and each link represents the number of drivers that neighboring subjects can handle and promote together; the greater the number, the wider the width of the link lines presented in the network. In the driver-driver network shown in Figure 4, each node represents a kind of driver, while each link represents a relationship between drivers. Te two 1-mode networks allow a clearer view of the key governing subjects and the KDs, and the strength of these interrelationships can be quantifed.

Centrality Analysis.
Degree centrality, intermediary centrality, and proximity centrality metrics are examined in conjunction with the recycling subject-driver network.
(1) Subject Centrality. Te results of the subject centrality analysis are shown in Table 3. Te table shows that S1 (government) has the highest centrality because it has the infuence to deal with the majority of drivers. S2 (owners), S3 (contractors), and S5 (recycling companies) are ranked second, third, and fourth, respectively, and the three are defned as core supply chain subjects in this paper. Tis fnding suggests that governments and core supply chain subjects have greater access to resources to control drivers and can generate greater impetus for collaborative governance, while at the same time, they are likely to work with a wider range of subjects. In terms of closeness and centrality, S1 (government) scores the highest, indicating that the government has the shortest path of infuence on other subjects, followed by S2 (owners), S3 (contractors), and S5 Liu et al. [12] I5: on-site inspection eforts for collection and classifcation

Inspections of C&DW collection and classifcation at construction sites by government functions
Yang and Song [17] I6: transportation regulatory legislation perfection degree Te government supervises and manages the perfection of the transport behavior regulations of transport enterprises Jia et al. [9] I7: transport route supervision eforts Government's eforts for regulation of transport companies' transport routes and exposure of violations Liu et al. [12] I8: completeness and rationality of the policy system Te completeness and rationality of the government's regulatory, binding, and supportive policies based on the characteristics of the new district development Liu et al. [12]; Zhao [13] I9: construction, operation, and maintenance of information platform Government information, interoperability platform, construction, perfection, and operation and maintenance management status Liu et al. [12] 6 Journal of Environmental and Public Health Te quality of recycled products produced by recycling companies Su and Si [14] I25: technology support for resource utilization process Relevant R&D units to dismantle, sort, resource treatment, recycled products production of the whole process, and support technology research and development status Akinade et al. [24] I26: fnancial support Financial support from fnancial institutions for recycling companies and R&D institutions Lin [25] I27: application ratio of recycled products Te percentage of construction and contractors using recycled products in new construction projects Yuan et al. [26] I28: publicity eforts for recycled products Government, enterprises, media, and other publicity eforts for recycled products and demonstration projects Wong [27] I29: attitude of stakeholders participated in governance Recycling stakeholders' attitudes and awareness of participation in collaborative governance Jain et al. [5]; Wong [27] Journal of Environmental and Public Health     Journal of Environmental and Public Health (recycling companies), indicating that the four subjects are most closely connected to other participating subjects and are in a more central position. Interestingly, the ranking of subjects in terms of degree, intermediation, and closeness centrality is similar. S6 (fnancial institutions) and S7 (public and media) have low degree centrality, intermediary centrality, and closeness centrality. Te main reason may be that these subjects have a low degree of expertise in the governance of recycling in new districts, especially the media and fnancial institutions. However, their contribution to collaborative recycling governance cannot be ignored. Specifically, the media can play an efective role in monitoring illegal dumping of the construction waste, assessing the strength of government regulation, and promoting benchmark projects, thus leading to macrolevel cooperation among subjects to collectively manage construction waste.
(2) Driver Centrality. Te results of drivers' centrality calculation are shown in Table 4, and the correlation analysis is as follows: Te factors with the highest degree of centrality are I8 (completeness and rationality of the policy system), I12 (new project review and approval system), followed by I3 (degree of collaboration between demolition and new construction plans), and I13 (tender document requirements for waste disposal), and most subjects must pay attention to these four KDs because they cannot be handled and promoted by only one subject. Tis shows that in order to achieve collaborative governance, while emphasizing a complete and reasonable incentive and constraint policy by the new regional government, active collaboration and cooperation of all supply chain subjects are needed to create a multisubject collaborative efect. Te study also found that the degree centrality of drivers such as I1 (special plan for recycling), I9 (construction, operation, and maintenance of an information platform), I16 (construction site management level), I20 (rationality of resource-based facility construction), and I23 (matching degree of recycled products and new district demand) are also high, indicating that these are also the tasks that collaborative governance should focus on.
Te drivers with the highest intermediary centrality are I8 (completeness and rationality of the policy system) and I12 (review and approval system for new projects), followed by I23 (matching degree of recycled products and new district demand) and I3 (degree of collaboration between demolition and new construction plans), which are similar to the elements with the highest degree of centrality.
Te higher the value of the intermediary centrality of drivers, the higher the number of times they act as a bridge for the shortest infuence path between nodes in the network, and the more critical their position in the network. Handling such drivers requires the cooperation and collaboration of more related subjects, and if these factors are not handled well or neglected, they can easily lead to other problems.
Te higher the value of intermediary centrality of drivers, the higher the number of times they act as a bridge of the shortest infuence paths between nodes in the network, and more critical their position in the network. Handling such drivers requires the cooperation and collaboration of more related subjects, and if these factors are not handled well or neglected, they can easily lead to other problems. For example, if the on-site management level cannot be improved through institutional constraints on the contractor, even if a feasible recycling special plan is formulated, and the coordination of demolition and new construction plans is strengthened, it will not be able to efciently match the output of recycled products with the demand of the new district. Te governance elements with the highest degree and betweenness centrality are I8 (completeness and rationality of the policy system) and I12 (review and approval system for new projects), followed by I3 (degree of collaboration between demolition and new construction plans) and I13 (tender document requirements for waste disposal). Tese factors are the same as the factors with the highest degree and betweenness centrality, and the factors with high closeness centrality are usually related to key participants and have a huge impact on the network.

Core-Periphery Analysis.
In order to further clarify the main subjects and KDs of recycling, this paper conducts the core-periphery analysis of the subject-driver network, and the density matrix is shown in Table 5. Te fnal ft is 0.772, indicating that the governance subject-driver network structure conforms to the ideal core-periphery structure. Te density of core to periphery and peripheral to core is 0.648 and 0.404, respectively, which indicates that the core governance subject is strongly correlated with edge drivers, but  the connection between the edge governance subject and KDs is relatively loose. Although fnancial institutions, media, and other governance subjects are closely related to drivers, the network of governance subject-driver presents a core-periphery structure with a high density of connection between cores. Based on this analysis, this paper identifes 4 core governance subjects and 15 KDs.

Discussion.
Based on the connotation of collaborative governance and the results of the social network analysis mentioned above, the strategies of collaborative governance are discussed as follows.

Governance Strategy Based on Organizational Collaboration
(1) Build a government-enterprise collaborative governance platform, and defne the power, responsibility, and obligation of each party based on the interest balance system. From Figure 3, it can be seen that the government can handle the most drivers, followed by owners, contractors, and recycling companies, showing a strong power similarity. Tis requires the government to maintain a leading position and perform government functions while paying attention to the core subjects of the supply chain, which has a promotion efect on recycling. Te government should build a collaborative governance platform to determine and fulfll their respective powers, responsibilities, and obligations in a complete, practical, and reasonable manner based on the balance of interests. (2) Build a beneft and risk distribution mechanism for cooperation and competition among supply chain subjects. Based on the strong similarity of power among owners, contractors, and recycling enterprises as shown in Figure 3, the three are typical cooperative and competitive coexistence relationships. Te power and responsibilities should be reasonably divided based on their respective positions and roles in the supply chain, and a win-win beneft and risk distribution mechanism should be constructed.
(3) Establish a coordination mechanism for the division of labor among relevant government departments. Based on the fact that I3 (degree of collaboration between demolition and new construction plans) and I2 (functional department's collaborative eforts) are in the top 50% of the ranking KDs, it indicates that the new district government should strengthen departmental linkage, realize information sharing, and establish a sound collaborative work mechanism for the whole process of recycling management. (4) Establish a media-public collaborative participation monitoring mechanism. Based on the fact that the media and public play an important role in advancing the top 50% of KDs, such as I27 (application ratio of recycled products) and I29 (attitude of stakeholders participated in governance), they should also be given the responsibility of social monitoring and promotion.

Governance Strategy Based on Resource Coordination
(1) Build a resource collaboration mechanism between the government and supply chain subjects. I8 (completeness and rationality of the policy system) is ranked No. 1 KD, which is closely related to the supply chain subjects; the government should systematically formulate a practical policy system based on the principle of win-win cooperation with relevant subjects on the basis of equal communication, and promote the division of labor and cooperation among supply chain subjects to carry out their work. At the same time, I12 (review and approval system for new projects) can be used as a breakthrough, and I3 (degree of collaboration between demolition and new projects) can be used as a means to ensure that recycling is implemented on the ground in a timely manner, and also to provide conditions to ensure the balance of construction waste supply and demand as much as possible. In order to guarantee the orderly implementation of recycling, the government should also compile a special plan for recycling (I1) and build an information platform (I9) on the basis of full research and extensive consultation, so as to guide the work of supply chain subjects in the new district from the top-level design and realize information sharing and Table 5: Core-periphery distribution of main subjects and drivers.   I1 I2 I3 I18 I19 I20 I8 I21 I16 I13 I27 I12 I9 I23 I19 I22 I4 I5 I7 I10 I11 I14 I17 I24 I25 I26 I28 I29 I15  S1 1 1 1 1  1  1 1 1 1  1  1  1 1 1  1  1 1 1 1 1  1  1  1  1  1  1  1  1  1  S2 1 1 1 1  1  1 1 1 1  1  1  1 1 1  1  1 1 1 0 1  0  1  1  1  1  0  1  1  0  S3 1 1 1 1  0  0 1 1 1  1  1  1 1 1  0  0 1 1 1 0  0  1  1  1  0  0  1  0  1  Journal of Environmental and Public Health communication between the government and the main subjects of the supply chain. At the same time, the supply chain subjects should also regulate and reasonably allocate their resources according to the requirements of policies and systems, for instance, owners should put forward the requirements for waste disposal in the tender documents according to policies (I13), contractors should manage the construction sites according to system requirements (I16), and recycling enterprises should reasonably build resource recycling facilities according to the recycling special plan (I20). (2) Build a resource collaboration mechanism for the main subjects of the supply chain. I13 (tender documents' requirements for waste disposal) and the application of recycled products in the bidding documents have an interactive relationship with I16 (construction site management level), I18 (C&DW quality), I27 (application ratio of recycled products), I20 (rationality of recycling facility construction), and I23 (matching degree of recycled products and new district demand). Terefore, on the premise of complying with the policy requirements of the new district, owners should conduct sufcient research and in-depth communication with relevant market subjects, and reasonably determine the requirements for construction waste treatment and application of recycled products in the bidding documents based on the principle of win-win cooperation and clarify the matching quotation requirements, payment methods, and division of risk and responsibility. Similarly, the resource application of contractors has an interactive relationship with the resource application of recycling enterprises. Tus, the supply chain subjects should actively participate in I9 (construction, operation, and maintenance of the information platform) and establish a resource collaboration mechanism for supply chain subjects to achieve win-win cooperation through information sharing on the information platform.

Conclusion
Tis paper analyzes the complex relationship between the subjects and drivers of recycling governance by using SNA in order to present the necessity of collaboration among multiple subjects and to fnd collaborative governance strategies. Trough literature analysis and feld research, 8 governance subjects and 29 drivers of recycling in the new district are identifed. Te analysis of subject centrality shows that the government has the highest degree of centrality, followed by owners, resource utilization companies, and contractors (including demolition companies), indicating that they have more infuence in the collaborative governance of recycling. Factor centrality analysis shows that drivers I8 (completeness and rationality of the policy system) and I12 (review and approval system for new projects) have the highest degree of centrality, followed by I3 (degree of collaboration between demolition and new construction plans), I13 (tender documents' requirements for waste disposal), I1 (special plan for recycling), I9 (construction, operation, and maintenance of the information platform), and I16 (Construction site management level). Te core-periphery analysis identifes a total of 4 core subjects and 15 KDs with strong relationships with each other. Finally, this paper proposes a collaborative strategy based on the organization of the subjects and resources under the background that diferent governing subjects promote and deal with multiple drivers through the analysis of the subject-subject network and driver-driver network. As a fundamental study on collaborative governance of recycling, this paper elucidates the complex relationship between governance subjects and drivers more systematically than previous studies by using 2 models of the social network analysis. Te research results provide a basis for each subject to clarify the key subjects that need to be collaborated and the KDs that need to be jointly promoted. Te proposed synergistic strategy provides a reference for the optimal allocation of responsibilities, rights, and resources of the management subjects, which can promote the development of the collaborative management of new district recycling, and thus realize the efcient recycling of construction waste during the new district development.

Data Availability
No data were used to support the fndings of this study.

Conflicts of Interest
Te authors declare that they have no conficts of interest.